Identification of genes associated with castration‑resistant prostate cancer by gene expression profile analysis

Abstract

Prostate cancer (CaP) is a serious and common genital tumor. Generally, men with metastatic CaP can easily develop castration‑resistant prostate cancer (CRPC). However, the pathogenesis and tumorigenic pathways of CRPC remain to be elucidated. The present study performed a comprehensive analysis on the gene expression profile of CRPC in order to determine the pathogenesis and tumorigenic of CRPC. The GSE33316 microarray, which consisted of 5 non‑castrated samples and 5 castrated samples, was downloaded from the gene expression omnibus database. Subsequently, 201 upregulated and 161 downregulated differentially expressed genes (DEGs) were identified using the limma package in R and those genes were classified and annotated by plugin Mcode of Cytoscape. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using Database for Annotation, Visualization and Integrated Discovery and KEGG Orthology Based Annotation System 2.0 online tools to investigate the function of different gene modules. The BiNGO tool was used to visualize the level of enriched GO terms. Protein‑protein interaction network was constructed using STRING and analyzed with Cytoscape. In conclusion, the present study determined that aldo‑keto reductase 3, cyclin B2, regulator of G protein signaling 2, nuclear factor of activated T‑cells and protein kinase C a may have important roles in the development of CRPC.

Figure 1.

Hierarchical clustering heat maps of the DEGs. The horizontal axis represents sample names, with GSM823845, GSM823846, GSM823847, GSM823851 and GSM823852 being the castration samples, and GSM823844, GSM823848, GSM823849, GSM823850 and GSM823853 the non-castration samples. The vertical axis indicates the clusters of DEGs: Red color stands for an expression level above the mean and green color stands for the expression lower than the mean. DEGs, differentially expressed genes.

Figure 2.

Top 30 hub genes in from the constructed protein-protein interaction network. The horizontal axis represents degree. The vertical axis indicated hub genes.

Figure 3.

Protein-protein interaction network constructed from the DEGs. Red circles indicate upregulated DEGs and green circles indicate downregulated DEGs. DEGs, differentially expressed genes.

Figure 4.

Topology parameters of protein-protein interaction networks. (A) Average clustering coefficient distribution. (B) Closeness centrality. (C) Neighborhood connectivity distribution. (D) Node-degree distribution. (E) Shortest path length distribution. (F) Topological coefficients.

Figure 5.

Top 5 modules obtained from the protein-protein interaction network of the differentially expressed genes. The circular and triangular nodes represent proteins (triangular nodes indicate hub proteins) and the grey lines represent interactions. Red circles indicate upregulated genes and green circles indicate downregulated genes.

Figure 6.

Biological processes, cellular components and molecular functions of differentially expressed genes were annotated and classified using GO analysis. The horizontal axis indicates the names of GO terms. GO, gene ontology.

Figure 7.

Directed acyclic graph based on the enrichment degree of GO terms. Color depth represents the degree of GO terms enrichment. GO, gene ontology.